Resources Policy 80 (2023) 103266

Contents lists available at ScienceDirect

Resources Policy
journal homepage: www.elsevier.com/locate/resourpol

Evaluation of mining projects subjected to economic uncertainties using the

Monte Carlo simulation and the binomial tree method: Case study in a
phosphate mine in Egypt
Ahmed Kamel a, Mohamed Elwageeh a, b, Stefano Bonduà c, Mohamed Elkarmoty a, *
a
Cairo University, Faculty of Engineering, Department of Mining, Petroleum, and Metallurgical Engineering, 1 Gamaa Street, 12613, Giza, Egypt
b
British University in Egypt, Suez Desert Road, El Sherouk City, 11837, Cairo, Egypt
c
University of Bologna, Department of Civil, Chemical, Environmental and Materials Engineering, Via Terracini 28, 40131, Bologna, Italy

A R T I C L E I N F O A B S T R A C T

Keywords: Mining project evaluation is a very complex process and time-consuming task. Any mining project passes through
Risk analysis several study stages, like conceptual, preliminary feasibility, and feasibility studies, using a systematic and
Monte Carlo simulation standardized approach. These studies allow investors to evaluate the exploitations alternatives according to
DCF
economic and technical criteria to make a better decision. This research aims to study a model for investment risk
Phosphate economics
Binomial tree
analysis developed for the case study of “Abu Tartur mining project, Egypt.” Abu Tartur plateau is rich in
phosphate rock, turned into phosphoric acid and fertilizers. This study used site-specific data that has been taken
from the literature to build evaluation models using the Monte Carlo Simulation (MCS) and the Binomial De­
cision Tree (BDT) methods. The Discounted Cash Flow (DCF) is considered a benchmark method for all other
evaluation methods. The MCS method uses a set of parameters that, in some cases, will give a higher NPV of the
project and some others with lower values, but this depends on the probability of the input parameters. The
developed evaluation models allowed a specific range of values, with confidence intervals as the MCS model. The
advantage of using a probabilistic approach helps in the decisional phase, allowing for a more precise overview
of the variability of the final economic value of the mine. Deterministic methods, like the DCF method, offer a
solution that is limited to uncertainty analysis.

1. Introduction deterministic method, while the Monte Carlo Simulation (MCS) (Amico,
2003) and the Binomial Decision Tree (BDT) (Brandao et al., 2005) are
The mining industry is a very risky activity involving very high classified as probabilistic methods. The mining ventures can be esti­
capital investment. The decision-makers have to deal with many kinds of mated using the Net Present Value (NPV) method (Gardner, 2015).
risks, from geologic, economic, environmental, and political risks Nonetheless, inputs are unpredictable, and their risks should be identi­
(Dehghani, 2018). Among the others, the primary sources of uncertainty fied before estimating the NPV of ore reserves (Erdem, 2008). This kind
in the mining industry are: (i) the amount of overburden and reserve of calculation does not give any information about the probability of
estimation; (ii) ore grade estimation; (iii) capital and operating costs; occurrence of the estimated NPV.
(iv) selling prices, (v) recovery ratio; (vi) dilution; (vii) environmental Uncertainty and risk management are critical for making decisions in
issues. Each parameter should be evaluated and determined by its effects the extraction of mineral resources in a sustainable manner. Uncertainty
on the project, as investigated by (Topal, 2008) (Dehghani and in mine planning can be evaluated, for example, with a geostatistical
Ataee-Pour, 2013) (Kopacz et al., 2019). simulation of the variables influencing the decision process, which
Each risk should be evaluated, estimated, and determined by its ef­ generates a series of equally likely models. Then, for each model, a
fects on the project and community. The Discounted Cash Flow - Rate of transfer function is applied. The outcome is a response function distri­
Return (DCF-ROR) (Trigeorgis, 2000) is the most used project evalua­ bution that determines the likelihood (or risk) that a particular area
tion method in mining project evaluation. It is possible to classify it as a meets the product quality requirements. High-risk operations frequently

* Corresponding author.
E-mail address: [email protected] (M. Elkarmoty).

https://doi.org/10.1016/j.resourpol.2022.103266
Received 27 December 2021; Received in revised form 15 December 2022; Accepted 19 December 2022
Available online 31 December 2022
0301-4207/© 2022 Elsevier Ltd. All rights reserved.
A. Kamel et al. Resources Policy 80 (2023) 103266

fail to meet ore quality requirements, lowering the value of mineral The duration of the mining project is set in all situations, and the tests
resources (Kloeckner et al., 2021). provide us with the histogram of potential consequences and the esti­
The preliminary investigation stage of any project is known as the mated benefit. The fundamental distinction resides in how money’s time
pre-feasibility study. The pre-feasibility study is performed to identify, worth is handled. The discount rate in the deterministic DCF-ROR is
assess, and choose the best business scenarios technically and finan­ called in MCS the “risk-free rate,” and in the option pricing method is
cially. In this early stage, shortcut methods are permissible and can be called “risk-neutral odds of options.” The most common method to
applied to determine small investment components and production measure the uncertainty is the sensitivity analysis, while the Monte
costs. If the chosen scenario is deemed viable, it is advised that the Carlo Simulation is another way to do that with more consistency
research be extended to feasibility to gain a more in-depth under­ (Fontes, Marcelio et al., 2020). Mining project evaluation is a dynamic
standing of the project scenario. In estimating the project’s chances of process, so it needs new evaluation methods, like Real Options Valuation
success, a feasibility study considers all the project’s essential compo­ (ROV) (Shafiee et al., 2009).
nents, including economic, technical, legal, and scheduling difficulties. According to the previous works in mining project evaluation, we
It is an engineering study that uses test work and engineering analysis to found that most economic evaluation is applied to the mineral’s ores like
assess whether or not the project should proceed to the final engineering Iron, Copper, Gold, etc., which are more valuable, of course (Tleubergen
and construction stage. It is a “go/no-go” decision point, meaning that et al., 2017). At the same time, the phosphate ores and other stone ores
the answer is sometimes “no” The second difference between these two did not find their way into such models. In this paper, a comparison of
stages is that once a project has progressed to the detailed feasibility the DCF, MCS, and BDT methods is presented and applied to a real case
study stage, corporations have typically invested significant resources study. We analyze their similarities and differences critically from three
and built a professional reputation assuming that the project would be points of view: how they treat volatility in parameter values, such as
practical. Several elements influence whether a project is worthwhile, metal price and cost; how they integrate the time value of money; and
including the project’s cost and return on investment or if the enterprise how they require managerial flexibility. We demonstrate that, given
earned sufficient money or sales from customers (Dou et al., 2020). their apparent distinctions, they are simply different features of a gen­
The DCF-ROR evaluation method evaluates if a mineral resource can eral project evaluation system which has its most straightforward type,
be considered a reserve and, consequently, a mine. It needs a consid­ the deterministic base case scenario. It is worth mentioning that
erable study and works not only a complete assessment and feasibility modeling the natural world involves simplification and loss of accuracy.
studies of the deposit but also, needs a complete design and simulation From the general perspective, all three methods can be achieved by
of the mine. If the simulated parameters show promising and positive dealing with validation elements (Galli and Armstrong, 1999). We
results, it is possible to exploit the mineral deposit properly. The detailed emphasized this research as a case study of the (Abu Tartur plateau,
feasibility study should contain four main fields: economic, financial, Egypt) phosphate mining project.
environmental, and political studies. Our study will focus on the eco­ Among the others, the most important parameters in economic un­
nomic effects due to price volatility and other financial parameters. certainty are metal price uncertainty and operating cost uncertainty.
Political and environmental constraints exist in all countries, especially The NPV was calculated using the three methods (DCF, MCS, and BDT)
in Egypt, but they are out of the scope of this research. The economic applied to the Abu Tartur phosphate mining project in four scenarios: (1)
evaluation of a mining project can be divided into two main quantities, assuming certainty for both pricing and operating costs, (2) assuming
the potential earnings, and the investment costs. If the difference be­ uncertainty for the price, (3) assuming uncertainty for both price and
tween the two quantities is high enough, the project will be feasible. operational costs, and (4) assuming uncertainty for the price, opera­
From the above explanation, the cash flow analysis is defined as the sum tional costs, interest rate, inflation rate, taxes’ rate, grade, and recovery.
of net profit, depreciation, depletion, deferred deduction, and amorti­ When both pricing and operating cost uncertainties were considered, the
zation, or the sales revenue minus operating costs and income taxes. mine evaluation presented a higher NPV (Dehghani & Ataee-pour,
The Binomial Decision Tree (BDT), introduced firstly by (Cox et al., 2012). The evaluation method based on the multidimensional bino­
1979), becomes nowadays one of the most used techniques in economic mial tree method was developed by (Dehghani et al., 2014): the pyramid
evaluation, which presents a simple option valuation discrete-time technique. The NPV has been computed under economic uncertainties to
model. The basic economic principles of arbitrage methods for option determine the efficiency of the pyramid technique. Finally, the obtained
pricing are apparent in this setting. Its creation involves only basic results were compared to the outcomes of other evaluation methods,
mathematics, but it includes the well-known Black-Scholes model (Black such as the binomial tree. To the best knowledge of the authors, no
and Scholes, 1973) as a particular limiting case, historically obtained previous research performing a comparison between DCF, MCS, and
only from much more complicated methods. In several cases, the basic BDT applied to the evaluation of phosphate mining projects with the
concept is readily committing itself to generalization. Black and Scholes same four scenarios is present in the scientific literature. Therefore,
considered continuous trading and expected stock values to be scenario four is unique, where seven uncertain parameters have been
lognormal distribution. Their approach relied on some very complex studied. Software to model the binomial tree approach is rare, therefore
mathematics, while it was needed for using a much simpler method we developed the binomial lattice algorithm, and it has been imple­
(Dehghani et al., 2014). So, a simple option valuation discrete-time mented and coded in macros in the Excel © environment to build the
model is developed with basic mathematics. It gave rise to a simple model. It was a unique effort to handle the problem of the complicated
and efficient numerical method to evaluate a mining project efficiently lattice. The price lattice of this mining project has nearly 253 cells and
(Ganguly et al., 2021) (Azimi et al., 2013) (Ikhsani and Nainggolan, the cash flow and discounted cash flow lattices are as well. In addition,
2021) (Amini et al., 2015) (Dehghani and Ataee-pour, 2012). phosphate mining has lacked economic studies. Therefore, with this
MCS is a standard extension of the deterministic NPV base case since paper, we contribute a progression to the previous research in the field
it accounts for the possibility that variables are subjected to uncertainty. of mining project economic evaluation.
Generic statistical models classify the input parameters, such as regular, The paper discusses briefly the nature of phosphate ore chemical
lognormal, uniform, and triangular models. Although it is possible to composing and its existence in Egypt. Then the developed models have
make associations between variables, they are most generally viewed as explained in detail the different techniques of mining project evaluation.
separate. In our approach, we will use the MCS jointly with the Black Results and discussion results are presented, followed by the conclusion.
and Scholes time series models (Black and Scholes, 1973) to integrate
the association of successive parameter values (e.g., metal price). These 2. Phosphate ore in Egypt
expanded simulations of Monte Carlo are somewhat close to those used
for the pricing of European options (Kopacz et al., 2017) (Chen, 2011). Phosphates are essential elements for life on Earth (Mitra et al.,

2
A. Kamel et al. Resources Policy 80 (2023) 103266

2020). Phosphates are used in agriculture, industrial, and chemical establishes that the total reserve is more than 5 billion tons of phosphate
process as well. Phosphorous in phosphate form is one of the rock (EMRA, 2021).
macro-elements necessary for plant growth (Nitro­ The richest phosphate rock in Egypt has been discovered in Sebaiya
gen-Phosphorous-Potassium). Farmers have long been concerned with (Nile Valley), Safaga (Red Sea), as well as Abu Tartur (the Western
nitrogen fertilizers only for increasing their crops, but with the Desert). The higher the amount of organic matter in the soil, the faster
improvement of agriculture research, it has been established that the phosphate rock dissolves. Phosphate rock biological solubilization is
phosphorus plays a significant role in natural plant growth and therefore more environmentally friendly than acidulation. Phosphate rock injec­
increases productivity (Elwageeh, 2017). The growing number of people ted with bacteria and Mycrrohizae has proven to be a practical
across the globe implies the growth of the demand for phosphorous. approximation of using phosphate rock for continuous crop production
The world consumption of phosphate rocks is more than 150 million (Hellal et al., 2019).
tons per year, where the commercial grade is about 30% of P2O5 or Mine reclamation is the process of restoring mined land to an
higher, as stated (Abouzeid, 2008). Egypt was granted a considerable ecologically functional or economically viable state. Although mine
reserve of phosphates. The Egyptian phosphate ore reserves are usually reclamation occurs after mining is completed, mine reclamation activ­
of low grade, so it needs mineral processing and removing gangue ities are planned before a mine is permitted or started. Liquidation is the
minerals. process of converting assets into cash or cash equivalents by selling them
Egypt is regarded as one of the world’s leading phosphate rock ex­ on the open market. An asset is a value-added component in finance. In
porters. Nevertheless, by refining this treasure to highly evaluated items, our case study, the reclamation and liquidation costs are not considered
the need to add significant value to Egyptian mineral wealth must be in the NPV calculations, therefore it is recommended to reassess the
accomplished. The unprocessed Phosphate rock can be used directly in project implementing these costs.
agriculture as soil fertilizer, and the processed mineral can be used in
different applications, like pharmaceuticals, personal care products, and 3. Methods
polishing agents in toothpaste with calcium. According to (Elwageeh,
2017), Phosphate is composed of apatite and a group of calcium phos­ The mining project evaluation process involves uncertainty. The
phate minerals that is the primary source of phosphoric acid in phos­ economic evaluation of mining projects can be classified into four sce­
phate fertilizers. Egypt has three main areas of phosphate, which is the narios. The First scenario uses Parameters With Fixed Values (PWFV).
belt of phosphate Fig. 1 from the east red seacoast to the New Valley in These values are considered “trusted values” that have been known
the western desert. This belt of phosphate is almost of deposition origin. before from the first stages of the pre-feasibility and feasibility studies
Abu Tartur plateau lies 50 km west of Kharga City, the capital of the (Barnes, 1980). Therefore, the economic evaluation process will use the
New Valley Governorate in southwestern Egypt. The Abu Tartur plateau DCF method. The second scenario assumes one uncertain parameter,
is perfectly connected to the urban through a first-class network of a while the others are fixed as the first scenario which is the ore price. In
heavy-load asphaltic road. It takes 45 min to go from the airport to the this scenario, the representation model is the BDT model. The third
mine. The phosphate extracted from the Abu Tartur mine ranges from 24 scenario assumes two uncertain parameters, and the MCS method is the
to 31% P2O5. It is ideal for producing Triple Super Phosphate (TSP), most common method used to deal with this kind of problem. The fourth
Single Super Phosphate (SSP), Nitrogen Phosphorus and Potassium scenario of the evaluation process implements seven uncertain param­
(NPK), Di-Ammonium Phosphate (DAP), and Mono-Ammonium Phos­ eters to the evaluation model using the MCS method.
phate (MAP) fertilizers and direct application. It is as well ideal for The deterministic method, like the DCF-ROR, uses the input pa­
producing Phosphoric Acid after using special treatment. It also exhibits rameters as PWFV. The BDT method mainly depends on predicting
significantly low toxic components like As, Cd, Pb, Cr, and Hg. The area mineral price and operating cost parameters. Prediction means that we
of the Abu Tartur plateau is about 1200 km2, and a raw estimation know the present value of such parameters and need to know its value in

Fig. 1. Location map of phosphorus deposits belt in Egypt, (EMRA, 2021).

3
A. Kamel et al. Resources Policy 80 (2023) 103266

the future and, in turn, predict its effect on the NPV, the Internal Rate of 3.2. The discount cash flow method (DCF)
Return (IRR), and the Payback Period (PP). On the other side, the MCS
will also start from present values, but it will depend on the distribution The classical deterministic method of economic analysis of a mining
of such parameters according to the variable nature of the parameters. project is the Discounted Cash Flow (DCF), derived by (Fisher, 1930).
This makes turbulence of parameters like price (Dehghani and Bogda­ DCF is the most reliable and popular method with three main measures
novic, 2018) reverse the investment from profit to loss. So, the MCS NPV, IRR, and PP.
technique is beneficial for decision-makers. MCS has been used in this The DCF analysis is commonly used to estimate the discounted po­
study to evaluate the NPV value considering parameter uncertainty tential cash flows over the project life to get its present value. The PWFV
(Mohapatra, 2009). is applied, so uncertainty and risk are not represented. The economic
study’s main aim is to help investors understand the estimated present
3.1. The net present value approach (NPV) value of the potential profit and expenses relative to the investment’s
costs (Stermole and Stermole, 1999).
The NPV must be computed to assess the value of the mine. In DCF,
the most popular approach is NPV, BDT, and MCS. NPV is defined as the 3.3. Monte Carlo simulation method (MCS)
cash inflow discounted to the present minus the investment cost. Ac­
cording to Eq. (3), the NPV shows the initial capital cost and the dis­ The MCS method is designed to model the likelihood of various
counted cash flow. So, the NPV is defined as the difference between the outcomes in a procedure that cannot be easily forecasted due to random
present value of cash inflows and the present value of cash outflows over variables. It is a method used in statistical and forecasting models to
some time. NPV is used in capital budgeting and investment planning to study the effects of risk and uncertainty (Brandimarte, 2014). A random
analyze the profitability of investment projects. NPV is the result of value is chosen from a given range between min. and max. For each
calculations used to find today’s value of a future stream of payments input parameter. These values are involved in the equation of the NPV
and costs. Most decision-makers utilize a single-figure output to deter­ computation. A standard simulation in MCS calculates many solutions,
mine the project’s financial worth. However, the NPV should always be using different parameter values chosen each time randomly. When the
calculated based on the project’s financial scale. There are two types of simulation is completed, as indicated in Fig. 2, we have many model
expenditures Operating expenses (OPEX) and capital expenditures outputs, each dependent on random input values. These results are used
(CAPEX) that a company incurs on an ongoing basis. to characterize the probability of the model producing different out­
CAPEX is a company’s major, long-term expense, whereas OPEX is comes (Samis and Davis, 2014), (Gamba, 2005) (NASA, 2017).
day-to-day costs. Equipment, machines, and trucks are examples of
CAPEX. Overhead expenses such as wages, utilities, and property taxes 3.4. The Binomial Decision Tree method (BDT)
are all examples of OPEX. For tax purposes, capital expenditures cannot
be deducted from income, whereas operating expenses can (Chen, The binomial model is a well-known, discrete alternative time
2021). Companies should be cautious about investing in projects with developed by (Cox et al., 1979). The BDT is a graph structure that maps
high CAPEX and OPEX but only a moderately positive NPV. all possible metal price trajectories over time, as the model permits
To calculate the NPV, the analyst should discount the future cash calculating the NPV approach. The structure is made up of nodes and
flow in addition to the investment cost, as shown in Eq. (3). The Cash branches. Each node in each layer corresponds to a possible metal price
Flow is calculated through Eq. (1) and Eq. (2): at a specific point in time. Nodes are marked with both traversal prob­
abilities and metal prices. A convenient indexing scheme has the layer or
Xt = (Pt − Ct ) × Qt − Ft − Dt (1)
time level defined by j (number between 1 and n, number of layers or
time steps), and the nodes within each layer (possible metal prices) by i
{ CFt = Xt (1 − Tt ) + Dt if Xt > 0
(number between 1 and m, number of nodes in the layer). The node
CFt = Xt + Dt if Xt ≤ 0 } (2) count m for any given layer will range from j to twice the number of
nodes in the previous layer, depending on whether the tree is recom­
where. bining or not. Each branch or path in a binomial pricing tree represents a
potential transition from one node to another later in the tree and is
Xt: taxable income at time t associated with a likelihood and a ratio. Higher node connections
Pt: price at time t represent probabilistic value (P0) and up factor (u), whereas lower node
Ct: operating cost at time t branches represent probabilistic value (P0) and down factor (d). Fig. 3
Qt: tonnage of production at time t shows a schematic binomial tree with three measures on the first metal
Ft: fixed cost at time t price at time zero (P0). The factors up (u) and down (d) and the likeli­
Dt: the deprecation at time t hood of occurrence (Pr) were calculated using the following formulation
CFt: cash flow of the project (Dehghani, 2018)(Brandao et al., 2005).
Tt: tax rate at time t. √̅̅̅̅
u = eσ δ t (4)
∑
n
CF
NPV = − CI + (3) √̅̅̅̅ 1
t=1 (1 + i)t d = e− σ δt
= (5)
u
where.
(1 + r) − d
pr = (6)
u− d
NPV: is the net present value.
CI: is the capital investment. The volatility of the metal price (σ), the risk-free rate (r), and the
CF: is the cash flow. stepping time (δt) are the primary inputs. In evaluating the proposed
t: is the time in years. Abu Tartur area project, we select three different scenarios, assuming a
n: is the project lifetime. combination of deterministic and probabilistic parameters.
i: is the discount rate.

4
A. Kamel et al. Resources Policy 80 (2023) 103266

Fig. 2. The procedure of Monte Carlo Simulations from input to output (NASA, 2017).

3.5. Computation methods (6). It is the method to forecast future prices. The future operating cost
data is determined using regression analysis depending on the inflation
To evaluate the feasibility study of a hypothetical phosphate mine parameter. So, a new Cash Flow (CF) binomial tree is estimated
project, we implemented several Excel worksheets for the three pre­ depending on the previous lattice, annual estimated operating cost, and
sented methods, DCF, BDT, and MCS method. From a computational Eq. (1) and Eq. (2). Finally, the CF is discounted in another tree that is
point of view, the DCF method is quite simple and easy to implement. constructed. The DCF will be estimated using Eq. (7) and opposite the
The BDT algorithm complexity has been implemented using the Visual previous lattices (from right to left).
Basic for Applications (VBA) programming language in an Excel envi­
pr × DCFt+1,k + (1 − pr ) × DCFt+1,k+1
ronment. The MCS method has been applied to our scenario using the DCFt,k = CFt,k + (7)
(1 + r)
software (@Risk, Crystal ball).
In the following, we compute the pre-feasibility study using the
where k is the node number at time t.
deterministic and probabilistic methods, as presented in Table 1.
In computation method 3, the price and operating cost have been
In computation method 1, we assumed deterministic metal price and
assumed as uncertain parameters, while the other parameters are PWFV.
operating cost. In this way, we can estimate the DCF value of the mine.
For each parameter, a probability distribution is chosen depending on
The purpose of these calculations is to make investors aware of the ex­
historical data. A close fit to the distribution is supposed to result in good
pected present value of future cash flows compared to investment costs.
predictions. In distribution fitting, it is necessary to use a distribution
Eq. (1), Eq. (2), and Eq. (3) can be used to get the base case NPV.
that is well-suited to the data. The project evaluation is obtained by the
In computation method 2, it is considered that metal price is un­
MCS method.
certain using the Binomial Decision Tree (BDT). A binomial tree is
In computation method 4, all dynamic input parameters are
constructed depending on the historical data and Eq. (4), Eq. (5), and Eq.

5
A. Kamel et al. Resources Policy 80 (2023) 103266

Fig. 3. Three-time step binomial tree.

and the equipment.

Table 1
Computation methods assumptions.
4.1. Input data
Computation Comp. Variable’s type Variable’s uncertain
method number Method assumptions
The input data that have been used in this work are literature data
1 DCF deterministic None from (Elwageeh, 2017). In Table 2, a resume of the technical parameter
2 CF, BDT deterministic and ore price
is reported. The Phosphate average grade (30%) and the cutoff grade
probabilistic
3 MCS deterministic and ore price and operating cost (25%) improve the company’s profit (Afify and Zaghloul, 2014) (Hellal
probabilistic et al., 2019). The stripping ratio is defined as the amount of overburden
4 MCS probabilistic ore price, operating cost, that should be removed to extract a given ore quantity. The hypothetical
grade, recovery, ROR, mine life is determined by dividing the reserve tonnage at the cutoff
inflation rate, and taxes rate
(million tons) (Azimi et al., 2013) over the ore production rate planned
(million tons/year), which gives 22 years for the production rate of 3.5
uncertain, including (price, operating cost, grade, recovery, interest million tons/year. Table 3 is resumed the financial parameters used in
rate, inflation rate, and Taxes rate), while the other parameters are this study.
PWFV. After the global economic crisis in 2009 and the revolutions that
occurred in Egypt in 2011 and 2016; Egypt has floated its currency in a 4.2. Output data
move that has reduced its value by almost 50% against the USD dollar in
2016. The Egyptian Central Bank considered this move as one of a list of For computation method 1, considering the deterministic metal price
reforms designed to strengthen confidence in the economy. The Central and operating cost, Fig. 4 represents the relation between NPV and DCF-
Bank has also increased interest rates by 3% reaching 14.75%. There­ ROR. The figure indicates that the NPV decreases with an increase in the
fore, we aimed to study the effect of variability of the interest rate by ROR. We notice that, at DCF-ROR = 0%, the NPV is $9.5 Billion, which is
using a probability distribution. an impossible case, of course. While at DCF-ROR 24%, NPV vanishes (i.
The proposed method of project estimation that consider the variable e., NPV = $0), and in this case, the DCF-ROR is called the IRR. By
uncertainty is the MCS. To perform the MCS, the first step is to develop increasing the ROR more than the IRR limit (24%), we obtain a negative
an analytical model to be evaluated. The second step is to generate a NPV which means that a risk case benefits the mining project.
probability distribution for each variable of the model, using subjective The NPV, IRR, and payback period are presented in Table 4. The
or historical data. The MCS model computes the outcomes of the project NPV, calculated at a 3% inflation rate and 10% rate of return, is esti­
using the statistical distribution of all the parameters in the NPV equa­ mated at $2.25 billion. NPV is a positive value which is a good indication
tion. The statistical distributions, such as normal, lognormal, triangular, for the project. The second measure is the IRR; in our case, we get nearly
and uniform, are used to evaluate the project considering the uncer­ 24% greater than the introductory rate ROR = 10%. It is also a good
tainty in parameters. We used the statistical distributions of the vari­
ables according to the available data and we assumed the statistical Table 2
distribution of parameters for variables with poor information. In every Technical parameters related to mine.
simulation, the values are selected randomly from each parameter dis­ Technical Parameters
tribution for every period and substituted into the NPV equation to
Average Grade 30.0%
generate one possible outcome of the project. We noted that the price Cutoff Grade 25.0%
and operating cost had identical distributions and values as in scenario Reserve Level at Cutoff (million tons) 77.50
3. Finally, many model outcomes are generated, and a corresponding Contained Value (million tons of P₂O₅) 23.25
model of probabilities vs. model outcomes is generated. Stripping Ratio 8.5
Ore Production Rate (t/d) 11,667
Mill Recovery 95.0%
4. Results Operating days/year 300
Mine Life (year) 22
This study investigates the uncertainty of prices and costs of the ore Ore Production Rate planned (million tons/year) 3.50

6
A. Kamel et al. Resources Policy 80 (2023) 103266

Table 3 When the input random variables are considered independent, the un­
Financial parameters. certainty can be underestimated or overestimated. In the case of a
Financial Parameters positive correlation, the value for each should be relatively high in one
simulation iteration and relatively low in next simulation. For negatively
Average (HPO4) Price ($/t) 766.0
Mine Operating Cost ($/t) 15.0 correlated inputs, one variable should be at the high end of the possible
Mill Operating Cost ($/t) 45.0 range for a given iteration, while the other should be at the low end. In
Total Operating Cost ($/t) 60.0 most cases, the variables are assumed to be independent to simplify the
Mine Capital Cost ($) 580,000,000 calculation. In reality, the majority of the variables are frequently
Mill Capital Cost ($) 800,000,000
Total Capital Cost ($) 1,380,000,000
correlated. In mining, for example, ore grades are positively correlated
Working Capital ($) 60,000,000 with ore recovery. In addition, it is shown that the prices and operating
Capitalized Exploration Cost ($) 1,000,000 costs are correlated with the inflation rate.
Depletion Allowance (%) 15.0% Due to the character of this preliminary feasibility study of the mine
Royalty (% Net Smelter Return) 5.0%
project, the statistical distribution of variables has been fitted to the
Income Tax Rate (%) 46.0%
Salvage Value (% of Capital Costs) 10.0% histogram distribution. The mean and the standard deviation have been
Real Risk-adjusted Discount Rate (ROR) (%) 10.0% used to define the variable distribution in the Crystal Ball software. The
Inflation (%) 3.0% mean values are corresponding to values used in scenario 1, for all the
parameters. The statistics of data have been estimated using SPSS soft­
ware (Farahani and Bayazidi, 2018) from the historical data fluctuation
indication and confidence in our progress in the project’s evaluation
for those parameters (AFA, 2021). Depending on the characteristics of
process. The payback period is estimated at five years, less than a
the phenomenon and the distribution, some probability distributions
quarter of the project’s lifetime (22 years). So, it is a very positive
can be fitted more closely to the observed frequency of the data rather
duration for payback.
than others. The distributions of the discount rate, income tax, and
Computation method 2 (uncertain metal price) is used to predict the
inflation parameters are taken based on the official websites of the
mineral price and calculate the free cash flow, the DCF, and the NPV.
Central Bank of Egypt (2022a), the Ministry of Finance of Egypt
Each of them has its decision tree and equations. So, in this scenario, one
(2022b), and the Central Agency for Public Mobilization and Statistics
of the input data is of uncertain value while the others are PWFV; this
(Egypt, 2022c). These parameters are distributed according to the
parameter is the ore price.
guidance values published on such websites, and they were considered
Table 5 contains the input data for each decision tree described
unrelated.
previously. The first price in the price tree is 766 $/ton, which is used in
the model of DCF. Volatility (σ) represents the change rate of historical
price data. In the BDT method, the risk-free rate of return is the theo­ Table 4
retical rate of return of an investment with zero risk. In a risk-free in­ Results from the DCF-ROR method.
vestment, the risk-free rate is the interest an investor would incur over a Results
specified period of time. The so-called “real” risk-free rate can be
Average NPV ($) 2,258,968,007
calculated by subtracting the current inflation rate from the yield of the IRR (%) 24.00
treasury bond matching the investment duration. Payback Period (years) 5.0
The parameters for the computation method 3 are reported in
Table 6 and Fig. 5. It assumes a distribution function for parameters
subject to uncertainty. Table 6 reports the basic statistics and the his­ Table 5
togram for the seven parameters used in calculations. The historical data Input parameters related to price for the BDT method and BDT-NPV result.
for the seven parameters have been analyzed and the statistical distri­
Input data
bution has been assumed using the best-fitting theoretical distribution.
The model of MCS generates random numbers from the range (0–100%) Volatility (σ) 28.5%
Up (u) 1.33
for all parameters of the theoretical statistical distribution, then applied
Down (d) 0.75
in Eq. (1), Eq. (2), and Eq. (3) to get the NPV value. This process has been Risk-free rate (rf) 7%
repeated 10000 times and finally, was possible to generate the proba­ Stepping time (t) 1.0 year
bility distribution of the NPV. Probability (Pr) 55%
First price in the Binomial lattice (Po) 766$/ton
Table 7 shows the correlation matrix between the input variables
Average NPV $2,126,849,233
which is an important factor to consider in Monte Carlo simulation.

Fig. 4. The relation between net present values (NPV) versus DCF-ROR.

7
A. Kamel et al. Resources Policy 80 (2023) 103266

Table 6
Input data about phosphoric acid in the MCS technique.
Parameter Distribution Mean Standard Deviation Skewness Kurtosis Histogram of the parameter

Average Price Normal 766.5 $/ton 16.5 $/ton 0.186 − 0.521

Operating Cost Normal 63 $/ton 2 $/ton 0.182 − 0.207

Average Grade Normal 29.4% 8% 0.121 − 0.756

Interest Rate Log-Normal 11% 3% 1.315 0.619

Inflation Log-Normal 10.4% 6% 0.895 0.143

Income Tax Rate Uniform 44.55% 0% – –

Mill Recovery Normal 95% 0% – –

8
A. Kamel et al. Resources Policy 80 (2023) 103266

Fig. 5. Statistical distribution of NPV and its parameters for scenario 3.

Table 7
Correlation index between input data in the MCS technique.
Parameter Average Price Operating Cost Average Grade Interest Rate Inflation Rate Income Tax Rate Mill Recovery

Average Price 1.0 – – – – – –

Operating Cost 0.77 1.0 – – – – –
Average Grade 0.80 0.50 1.0 – – – –
Interest Rate − 0.45 − 0.22 0.03 1.0 – – –
Inflation Rate 0.94 0.88 0.28 0.80 1.0 – –
Income Tax Rate 0.66 − 0.40 − 0.08 0.01 0.01 1.0 –
Mill Recovery 0.30 0.60 0.05 0.01 0.01 0.01 1.0

Fig. 5 represents the results of the computation method 3. It reports Computation method 4 is represented in Table 6 and Fig. 6. It con­
the NPV as a function of the parameter’s probability. It indicates that the siders the case in which most input parameters are subject to uncer­
mean value of MCS is near the base case, under the effect of uncertainty tainty. As shown in Table 6, we have seven parameters under the effect
in the parameters (price and operating cost). The graph reveals that for of uncertainty. Fig. 6 shows the results of the MCS run obtained using
the uncertainty of the input parameter of 50% (Median value), the ex­ 10000 iterations. The graph reveals that for the uncertainty of the input
pected value of the NPV was 2.35 billion dollars. If we consider a con­ parameter of 50% (Median value), the expected value of the NPV was
fidence interval of 95%, the expected value of the NPV will be between 3.04 billion dollars. If we consider a confidence interval of 95%, the
(0.00–4.26) billion dollars. expected value of the NPV will be between (0.00–5.94) billion dollars. In

Fig. 6. Statistical distribution of NPV and its parameters for scenario 4.

9
A. Kamel et al. Resources Policy 80 (2023) 103266

MCS, assuming that the ore price, operating cost, grade, and other pa­ (NPV>0) for the feasibility of the project. The IRR was 24%, which is an
rameters are subjected to uncertainty, the NPV is calculated considering encouraging result as well. Encouraging results for mining projects are
uncertainty and offers more realistic values, than the other methods. As considered when IRR > ROR10%. Finally, the computed payback period
shown in Fig. 6, the distribution of the output of the NPV is lognormally was five years. So, it is also fine because the Payback Period is less than
skewed to the left. This effect comes from considering the lognormal the mine life, which was established in 22 years (as in Table 2).
distribution of some variables used in the simulation. The second scenario has been computed using the BDT method. The
average value for this scenario was NPV = $2.13 billion, which is lower
5. Discussion than the reference case of the DCF by − 6%.
The MCS technique for the third computation method was obtained
The DCF method deals with technical and financial parameters in a considering the uncertain mineral price and operating cost parameters.
deterministic manner. BDT technique handles the parameters uncer­ Results show that the NPV was higher than the reference scenario by 4%,
tainty with a discrete probability for each node, but the tree becomes which is a good evaluation result of the method.
exponentially larger with increasing the time step compared to the mine The fourth scenario worked for seven uncertain parameters using the
life. The MCS was better than the DCF and the BDT methods of handling MCS method. Results show that the computed average of the computed
the uncertainty using continuous distribution for technical and financial NPVs was higher than the reference scenario by 34%. The confidence
parameters. interval was higher in scenarios 3 and 4.
The DCF method assumes that the scenario used PWFV, so the These project evaluation results were hopeful and can encourage the
decision-maker cannot have a clear overview of the final results of the investor to consider the mining project with more confidence. Fluctua­
parameter change. The BDT method deals with different strategies and tions of the ore price and the cost are not represented in the determin­
shows all the probable outcomes for the NPV obtained from these stra­ istic method, while in the probabilistic methods they are taken into
tegies. The MCS method focuses on modeling the uncertainties and ne­ consideration. Therefore, the MCS method can study thousands of trials
glects managerial flexibility, but it shows the distribution of the NPV or scenarios that may happen in the real world and give the probability
before starting the project. of the results and their confidence intervals value. The MCS can consider
MCS is a natural extension for the base case of NPV since the vari­ the real-world parameter variability, in order to compute thousands of
ables are not known with certainty. These variables are described using scenarios and compute the probability of the obtained NPV. The issue
statistical distributions such as normal, log normal, and uniform distri­ here is how to represent the parameters well so that we can evaluate the
bution. Although it is allowed to make correlations between variables, project accurately. In our case study, we found that the project is prof­
they are viewed as independent. In MCS, the project life is fixed in all itable from the DCF calculations, but what if the price decreases, and
cases, and the results have the histogram of future results and the ex­ what if the cost increases? The sensitivity analysis or what-if analysis
pected value. The underlying distinction lies in how the time value of can answer these questions for limited cases and limited parameters. The
money is treated. In the BDT method, the maximum expected value is MCS aims to consider these problems, therefore we can represent the
estimated by folding the tree back from the outer end branches toward fluctuation of multi-parameters in one model and at any stage of the
the start. They have some characteristics such as: (i) how they handle the project.
time value of money (discount rate vs. risk-free rate plus change of
probability), (ii) how they allow for uncertainty in parameter values, 6. Conclusion
and (iii) whether they incorporate managerial flexibility.
Scenario 1 indicates that using the PWFV, the mine project is The work aims to assess phosphate mining project scenarios in the
feasible. Scenario 2 considers the variability of the mineral price, indi­ Abu-Tartur region, Egypt. Different methods have been used to estimate
cating that in some cases, the project is not feasible, and in others, the the NPV, IRR, and PP in the present work. The latter parameters have
NPV is greater than the NPV obtained by Scenario 1. Scenarios 3 and 4 been computed using different computation methods under various
allow for obtaining similar results, with the added value to give jointly scenarios. Discounted Cash Flow (DCF) is commonly used to evaluate
with the NPV a measure of the probability of success of the mining mining projects, while other methods like Binomial Decision Tree (BDT)
project. In addition, the MCS method allows re-generating during the and Monte Carlo Simulation (MCS) enable the decision-maker to eval­
mine exploitation by updating the random used values with the actual uate the project clearly and consider the risk in the project due to un­
values, allowing for back analysis. certainty in parameters. Finally, the BDT method results in a specific
Table 8 shows the NPV obtained using the four different computation value of NPV depending on the probability of input parameters. We can
methods. The reference case was the DCF method, which used the conclude that the MCS method is favorable compared to other methods.
deterministic value of the technical and financial parameters. These The MCS method applies a set of parameters that, in some situations,
parameters mainly come from the feasibility and pre-feasibility studies produce better results than others, although this is dependent on the
made on the project. According to the computation of the NPV, it was likelihood of the input parameters. It is the best way since it aids in
estimated that NPV was $2.26 billion, which is an encouraging value considering the uncertainty of the parameters and variability assessment
and determining which parameters greater influence the outcomes.
Table 8 In order to consider uncertainty, the distributions of the input vari­
Summary of the results. ables were tested for goodness of fit with the observed data using his­
Case No. Average Change of Confidence Notes tograms. The parameters of the normal, lognormal, and uniform
NPV (B $) the base Interval distributions, i.e., the mean and standard deviation were also calculated
case from the data for each variable. The MCS for project evaluation over­
Scenario 2.26 0% 0% DCF (Base Case) plays the importance of the choice of which distribution to use to
1 represent input variables and underplays the importance of assessing
Scenario 2.13 − 6% <50% BDT (uncertain and including correlations between the variables. Simulation runs of the
2 mineral price)
evaluation model including correlations showed that correlations must
Scenario 2.35 4% 95% (0–4.26 MCS (uncertain
3 B$) price and operating be included in Monte Carlo Simulation otherwise the analysis leads to an
cost) underestimate of risk. It is relatively more important to investigate and
Scenario 3.04 34% 95% (0–5.94 MCS (uncertain allow correlations between the variables than to select the best fit dis­
4 B$) multi-input tribution to represent the parameters in the MCS.
parameters)
The MCS technique allows to improve project estimation and

10
A. Kamel et al. Resources Policy 80 (2023) 103266

comprehension of the variability of the results. Real validation of the Egypt, 2022c. Central agency for public mobilization and statistics. https://www.
capmas.gov.eg/HomePage.aspx.
proposed method will be possible when the mining project will start.
Elwageeh, M.A., et al., 2017. Maximum Utilization of Egyptian Phosphorities. Road Map.
Future work can consider a comparison of the results obtained in this EMRA, 2021. Egyptian Ministry of Petroleum and Mineral Resources [WWW Document].
study with other project evaluation emerging methods, like the Real URL. http://emra.gov.eg.
option Valuation and Cash Flow at Risk (Sauvageau and Kumral, 2018), Erdem, Ö., 2008. Uncertainty Assessment for the Evaluation of Net Present Value of a
Mineral Deposit. Cell. The Graduate School of Applied and Natural Sciences of
as well as considering the liquidation/reclamation cost of the project Middle East Technical University. https://doi.org/10.1016/j.cell.2009.01.043.
and the mineral grade modeling using, for example, geostatistical Farahani, H., Bayazidi, S., 2018. Modeling the assessment of socio-economic and
methods. environmental impacts of sand mining on local communities: a case study of Villages
Tatao Riverbank in North-western part of Iran. Resour. Pol. 55, 87–95. https://doi.
org/10.1016/J.RESOURPOL.2017.11.001.
Data availability Fisher, I., 1930. The Theory of Interest. Sentry press, New york.
Fontes Marcelio, P., Koppe Jair, C., Albuquerque, N., 2020. Comparison between
traditional project appraisal methods and applied to mining planning uncertainty
Data will be made available on request. analysis. Int. J. Min. Miner. Eng. 73, 261–265.
Galli, A.G., Armstrong, M., 1999. Comparison of Three Methods for Evaluating Oil
References Projects. https://doi.org/10.2118/57894-MS.
Gamba, A.N., 2005. Real options valuation: a Monte Carlo approach. SSRN Electron. J.
https://doi.org/10.2139/ssrn.302613.
Abouzeid, A.M., 2008. Physical and thermal treatment of phosphate ores. An overview
Ganguly, S., Basu, R., Nagarajan, V., 2021. A Binomial Decision Tree to Manage Yield-
85, 59–84. https://doi.org/10.1016/j.minpro.2007.09.001.
Uncertainty in Multi-Round Academic Admissions Processes. https://doi.org/
AFA, 2021. Arab Fertilizer Association (AFA) [WWW Document]. http://arabfertilizer.
10.1002/nav.22012. WILEY 1–17.
org/.
Gardner, N.A., 2015. A Discount Cash Flow and Real Option Valuation Approach for Iron
Afify, R.R.M., Zaghloul, A.M., 2014. Chemical characterization of phosphate rock
Ore Mining Company. McGill University, Montreal.
applied in arid region. Middle East Journal of Agriculture Research 3 (1), 59–70,
Hellal, F., El-Sayed, S., Zewainy, R., Amer, A., 2019. Importance of phosphate Rock
2014 3, 59–70.
application for sustaining agricultural production in Egypt. Bull. Natl. Res. Cent. 43
Amico, M., 2003. Simulation Methodology for Collateralized Debt and Real Options: A
https://doi.org/10.1186/s42269-019-0050-9.
New Methodology to Evaluate the Real Options of Investment Using Binomial Trees
Ikhsani, S., Nainggolan, Y., 2021. Real option valuation of coal mining project using
and Monte Carlo Simulation.
binomial lattice model. APIAR 7.
Amini, M., Alijani, F., Mozaffari, Z., 2015. Economic evaluation of mining projects under
Kloeckner, J., Alves, J.L.O., Silva, F.H.T., Guimaraes, O.R.A., Bassani, M.A.A., Costa, J.F.
conditions of uncertainty for prices and operating costs. Int. J. Innovat. Appl. Stud.
C.L., 2021. Application of risk assessment to improve sustainability in bauxite
10, 881–890.
mining. Resour. Pol. 74, 102328 https://doi.org/10.1016/J.
Azimi, Y., Osanloo, M., Esfahanipour, A., 2013. An uncertainty-based multi-criteria
RESOURPOL.2021.102328.
ranking system for open pit mining cutoff grade strategy selection. Resour. Pol. 38,
Kopacz, M., Galica, D., Dyczko, A., Jarosz, J., 2019. Economic valuation of coal deposits
212–223. https://doi.org/10.1016/j.resourpol.2013.01.004.
– the value of geological information in the resource recognition process. Resour.
Barnes, M.P., 1980. Computer-Assisted Mineral Appraisal and Feasibility. Society of
Pol. 63 https://doi.org/10.1016/j.resourpol.2019.101450.
Mining Engineers, New York.
Kopacz, M., Kryzia, D., Kryzia, K., 2017. Assessment of sustainable development of the
Black, F., Scholes, M., 1973. The pricing of options and corporate liabilities. J. Polit.
hard coal mining industry in Poland with the use of bootstrap sampling and copula-
Econ. 81, 637–657. https://doi.org/10.1086/260062.
based Monte Carlo simulation. J. Clean. Prod. 159, 359–373. https://doi.org/
Brandao, L., Dyer, J.S., Hahn, W.J., 2005. Using Binomial Decision Trees to Solve Real-
10.1016/j.jclepro.2017.05.038.
Option Valuation Problems. https://doi.org/10.1287/deca.1050.0040.
Mitra, D., Andelkovic, S., Panneerselvam, P., Senapati, A., Vasic, T., Ganeshamurthy, A.
Brandimarte, P., 2014. Handbook in Monte Carlo Simulation.
N., Chauhan, M., Uniyal, N., Mahakur, B., Radha, T.K., 2020. Phosphate-solubilizing
Chen, X., 2011. American option pricing formula for uncertain financial market. Int. J.
microbes and biocontrol agent for plant nutrition and protection: current
Oper. Res.
perspective. Commun. Soil Sci. Plant Anal. 1–13. https://doi.org/10.1080/
Chen, Y., 2021. Evaluating the influence of energy prices on tight oil supply with
00103624.2020.1729379, 00.
implications on the impacts of COVID-19. Resour. Pol. 73, 102129 https://doi.org/
Mohapatra, D., 2009. Investment Analysis in Mining Industry. National Institute of
10.1016/J.RESOURPOL.2021.102129.
Technology.
Cox, J.C., Ross, S.A., Rubinstein, M., 1979. Option pricing: a simplified approach.
NASA, 2017. Monte Carlo Simulation [WWW Document]. URL. https://www.nasa.
J. Financ. Econ. 7, 229–263. https://doi.org/10.1016/0304-405X(79)90015-1.
gov/centers/ivv/jstar/monte_carlo.html.
Dehghani, H., 2018. Evaluation of the mining projects under economic uncertainties
Samis, M., Davis, G.A., 2014. Using Monte Carlo simulation with DCF and real options
using real option valuation. In: International May Conference on Strategic
risk pricing techniques to analyse a mine financing proposal. Int. J. Financial Eng.
Management. Bor, Serbia, pp. 1–12.
Risk Management 1, 264. https://doi.org/10.1504/ijferm.2014.058765.
Dehghani, H., Ataee-Pour, M., 2013. Determination of the effect of economic
Sauvageau, M., Kumral, M., 2018. Cash flow at risk valuation of mining project using
uncertainties on mining project evaluation using Real Option Valuation. Int. J. Min.
Monte Carlo simulations with stochastic processes calibrated on historical data. Eng.
Miner. Eng. 4, 265–277. https://doi.org/10.1504/IJMME.2013.056850.
Econ. 63, 171–187. https://doi.org/10.1080/0013791X.2017.1413150.
Dehghani, H., Ataee-pour, M., 2012. Determination of the effect of operating cost
Shafiee, S., Topal, E., Nehring, M., 2009. Adjusted Real Option Valuation to Maximise
uncertainty on mining project evaluation. Resour. Pol. 37, 109–117. https://doi.org/
Mining Project Value - A Case Study Using Century Mine. Australasian Institute of
10.1016/j.resourpol.2011.11.001.
Mining and Metallurgy Publication Series, pp. 125–134.
Dehghani, H., Ataee-pour, M., Esfahanipour, A., 2014. Evaluation of the mining projects
Stermole, J.F., Stermole, J.M., 1999. Economic Evaluation and Investment Decision
under economic uncertainties using multidimensional binomial tree. Resour. Pol. 39,
Methods. Investment Evaluation Corporation, Golden, Colorado, Colorado.
124–133. https://doi.org/10.1016/j.resourpol.2014.01.003.
Tleubergen, M.A., Baimbetov, M.K., Mukanova, A.B., Kurmangazina, S.K.,
Dehghani, H., Bogdanovic, D., 2018. Copper price estimation using bat algorithm.
Satanbekov, N.A., Abayeva, N.E., Kabdualiyev, U.B., 2017. Evaluation technique of
Resour. Pol. 55, 55–61. https://doi.org/10.1016/j.resourpol.2017.10.015.
non-ferrous metal ore deposits. Espacios 38.
Dou, S. Quan, Liu, J. Yi, Xiao, J. Zhong, Pan, W., 2020. Economic feasibility valuing of
Topal, E., 2008. Evaluation of a mining project using discounted cash flow analysis,
deep mineral resources based on risk analysis: songtao manganese ore - China case
decision tree analysis, Monte Carlo simulation, and real options using an example.
study. Resour. Pol. 66, 101612 https://doi.org/10.1016/J.
Int. J. Min. Miner. Eng. 1, 62–76. https://doi.org/10.1504/IJMME.2008.020457.
RESOURPOL.2020.101612.
Trigeorgis, L., 2000. Managerial Flexibility and Strategy in Resource Allocation, fifth ed.
Egypt, 2022a. Central Bank of Egypt [WWW Document]. https://www.cbe.org.eg/ar
the MIT Press, Cambridge.
/Pages/default.aspx.
Egypt, 2022b. Ministry of finance [WWW document]. https://www.mof.gov.eg/en/.
https://www.mof.gov.eg/en/ (accessed 12.14.22).

11